Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Kimi K2.7 Code is generally available in GitHub Copilot

github.blog

Post reply on HN
Print view View original

31–40 of 201 posts

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#31

Post by rusticpenn »

asimovDev wrote: post #28
kingstnap wrote: post #26

Input: $0.95 Cache hit (most important): $0.19 Output: $4.00 This is the same as how much Moonshot charges for it, and it puts it at roughly the price of GPT 5.4 mini, not a bad option. For some context here is a stupid prompt that wastes tokens: "Play a game of tic tac toe against yourself on a 5x5 board, you need 5 in a row to win." It costs $0.006 on Kimi K2.7, and you get to see the whole raw reasoning trace. GPT…

when i will be extremely bored, I think I will make two models play chess against each other. I bet there's a chess benchmark / llm tournament already somewhere

Models are bad at chess. I am using a middleman to help models play chess and experimenting. https://abhay-ai.github.io/R_Daneel_AI/

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#32

Post by ubanholzer »

kingstnap wrote: post #26

Input: $0.95 Cache hit (most important): $0.19 Output: $4.00 This is the same as how much Moonshot charges for it, and it puts it at roughly the price of GPT 5.4 mini, not a bad option. For some context here is a stupid prompt that wastes tokens: "Play a game of tic tac toe against yourself on a 5x5 board, you need 5 in a row to win." It costs $0.006 on Kimi K2.7, and you get to see the whole raw reasoning trace. GPT…

Nice idea. I just asked Haiku to do the same in Claude Chat on iOS: it created a interactive react game, implemented the rules and let it play. Clever move for 1$ input and 5$ output, Anthropic!

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#33

Post by bel8 »

k__ wrote: post #27
arikrahman wrote: post #23

Earlier quoted context omitted.

I had a similar experience moving away from Copilot within Zed. Now using the reasonix harness for Deepseek that makes cache hits almost free. And that's with unsubsidized American providers like Digital Ocean or Cloudflare.

Nice. I paid $6 yesterday for DeepSeek V4 Flash on OpenRouter. That's like $120 dollar for a month, and it's not even a good model.

For DS4 it's much cheaper and reputable to use OpenCode Go $10/mo subscription, or directly with DeepSeek API.

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#34

Post by epolanski »

k__ wrote: post #27
arikrahman wrote: post #23

Earlier quoted context omitted.

I had a similar experience moving away from Copilot within Zed. Now using the reasonix harness for Deepseek that makes cache hits almost free. And that's with unsubsidized American providers like Digital Ocean or Cloudflare.

Nice. I paid $6 yesterday for DeepSeek V4 Flash on OpenRouter. That's like $120 dollar for a month, and it's not even a good model.

That's quite an achievement, I managed to spend only 2$ on 16 different tasks of v4 pro.

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#36

Post by Vinnl »

taspeotis wrote: post #29
nsoonhui wrote: post #21

I used GitHub Copilot for my VS 2026 development and switched between ChatGPT and Claude. That was before I discovered Claude Code and the Codex app. Copilot was OK for my purposes, and the USD 10 per month fee was enough for my usage. However, last month they introduced a new pricing model ( I know the old pricing was not sustainable), and my USD 10 was exhausted within days. Because of that, I switched to Claude Co…

The harness is super important, what tools are available and the system prompts vary from harness to harness. Anthropic seems to have a modest lead on their harness and models, so it’s a best-of-both-worlds scenario. > I'm not sure what Microsoft is doing behind the scenes It’s probably the exact same model, but the tools and the prompts around it are worse, so you get worse results.

So if you use Claude via Copilot in Zed... You use Zed's harness, I think? What does Copilot do, at that point?

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#38

Post by toyg »

arikrahman wrote: post #23
nsoonhui wrote: post #21

I used GitHub Copilot for my VS 2026 development and switched between ChatGPT and Claude. That was before I discovered Claude Code and the Codex app. Copilot was OK for my purposes, and the USD 10 per month fee was enough for my usage. However, last month they introduced a new pricing model ( I know the old pricing was not sustainable), and my USD 10 was exhausted within days. Because of that, I switched to Claude Co…

I had a similar experience moving away from Copilot within Zed. Now using the reasonix harness for Deepseek that makes cache hits almost free. And that's with unsubsidized American providers like Digital Ocean or Cloudflare.

I tried using Zed but with local models it constantly breaks on tool calls. I wanted to like it but the smell of vibing is just too much.

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#39

Post by w4yai »

andhuman wrote: post #4

Finally an alternative to the big dogs that a company can use. People have been asking for a way to run the Chinese models from a trusted provider. Here GitHub delivered! The performance, if we trust the benchmarks, put it at Sonnet 4.6. Let’s see if it’s worth it with GitHubs pricing.

> People have been asking for a way to run the Chinese models from a trusted provider

I'm going to be called a chiller again, but at this point I don't care as it is relevant. Synthetic runs their own models for a reasonable price, GLM5.2 & Kimi K2.7-Code included.

Referral link :

https://synthetic.new/?referral=kwjqga9QYoUgpZV

Re: Kimi K2.7 Code is generally available in GitHub Copilot

#40

Post by arcanemachiner »

toyg wrote: post #38
arikrahman wrote: post #23

Earlier quoted context omitted.

I had a similar experience moving away from Copilot within Zed. Now using the reasonix harness for Deepseek that makes cache hits almost free. And that's with unsubsidized American providers like Digital Ocean or Cloudflare.

I tried using Zed but with local models it constantly breaks on tool calls. I wanted to like it but the smell of vibing is just too much.

You using models released this year? I hear this complaint a lot, and it's often due to using an old model which is not as good at tool calling as newer models.
Post reply on HN